Spaces:
Sleeping
Sleeping
Update link4 (2).py
Browse files- link4 (2).py +2 -213
link4 (2).py
CHANGED
@@ -1,181 +1,3 @@
|
|
1 |
-
# -*- coding: utf-8 -*-
|
2 |
-
"""link4.ipynb
|
3 |
-
|
4 |
-
Automatically generated by Colab.
|
5 |
-
|
6 |
-
Original file is located at
|
7 |
-
https://colab.research.google.com/drive/1yTE900ZWoLy3vQwKE1Y-Qbm263XCIuN7
|
8 |
-
"""
|
9 |
-
|
10 |
-
!pip install selenium
|
11 |
-
!pip install webdriver-manager
|
12 |
-
!pip install pyshark
|
13 |
-
!pip install gradio
|
14 |
-
!apt-get update
|
15 |
-
!apt-get install -y tshark
|
16 |
-
!tshark --version
|
17 |
-
!pip install gradio requests scapy joblib pyshark
|
18 |
-
|
19 |
-
from google.colab import drive
|
20 |
-
drive.mount('/content/drive')
|
21 |
-
|
22 |
-
import pandas as pd
|
23 |
-
from sklearn.model_selection import train_test_split, cross_val_score
|
24 |
-
from sklearn.ensemble import ExtraTreesClassifier
|
25 |
-
from sklearn.metrics import classification_report
|
26 |
-
import joblib
|
27 |
-
import subprocess
|
28 |
-
import time
|
29 |
-
from selenium import webdriver
|
30 |
-
from selenium.webdriver.chrome.service import Service
|
31 |
-
from webdriver_manager.chrome import ChromeDriverManager
|
32 |
-
from selenium.webdriver.chrome.options import Options
|
33 |
-
import pyshark
|
34 |
-
import numpy as np
|
35 |
-
|
36 |
-
file_paths = [
|
37 |
-
'/content/drive/MyDrive/Colab Notebooks/link1/Friday-WorkingHours-Afternoon-DDos.pcap_ISCX.csv',
|
38 |
-
'/content/drive/MyDrive/Colab Notebooks/link1/Friday-WorkingHours-Afternoon-PortScan.pcap_ISCX.csv',
|
39 |
-
'/content/drive/MyDrive/Colab Notebooks/link1/Friday-WorkingHours-Morning.pcap_ISCX.csv',
|
40 |
-
'/content/drive/MyDrive/Colab Notebooks/link1/Monday-WorkingHours.pcap_ISCX.csv',
|
41 |
-
'/content/drive/MyDrive/Colab Notebooks/link1/Thursday-WorkingHours-Afternoon-Infilteration.pcap_ISCX.csv',
|
42 |
-
'/content/drive/MyDrive/Colab Notebooks/link1/Thursday-WorkingHours-Morning-WebAttacks.pcap_ISCX.csv',
|
43 |
-
'/content/drive/MyDrive/Colab Notebooks/link1/Tuesday-WorkingHours.pcap_ISCX.csv',
|
44 |
-
'/content/drive/MyDrive/Colab Notebooks/link1/Wednesday-workingHours.pcap_ISCX.csv'
|
45 |
-
]
|
46 |
-
|
47 |
-
# Combine all files into a single DataFrame
|
48 |
-
df = pd.concat([pd.read_csv(file) for file in file_paths], ignore_index=True)
|
49 |
-
|
50 |
-
# Strip any leading or trailing spaces from column names
|
51 |
-
df.columns = df.columns.str.strip()
|
52 |
-
|
53 |
-
# Print the first few rows and column names to verify
|
54 |
-
print("Columns in DataFrame:")
|
55 |
-
print(df.columns)
|
56 |
-
|
57 |
-
# Check if 'Label' exists
|
58 |
-
if 'Label' not in df.columns:
|
59 |
-
print("Error: 'Label' column not found in the dataset.")
|
60 |
-
else:
|
61 |
-
# Proceed with mapping the labels to "benign" or "malicious"
|
62 |
-
label_mapping = {
|
63 |
-
'BENIGN': 'benign',
|
64 |
-
'DDoS': 'malicious',
|
65 |
-
'PortScan': 'malicious',
|
66 |
-
'Bot': 'malicious',
|
67 |
-
'Infiltration': 'malicious',
|
68 |
-
'Web Attack': 'malicious',
|
69 |
-
# Add other malicious classes here if necessary
|
70 |
-
}
|
71 |
-
|
72 |
-
# Map the labels and fill missing values with 'malicious'
|
73 |
-
df['Label'] = df['Label'].map(label_mapping).fillna('malicious')
|
74 |
-
|
75 |
-
# Convert categorical labels to numerical
|
76 |
-
df['Label'] = df['Label'].astype('category').cat.codes
|
77 |
-
|
78 |
-
# Define features and target
|
79 |
-
all_features = df.columns.drop('Label')
|
80 |
-
features = df[all_features]
|
81 |
-
target = df['Label']
|
82 |
-
|
83 |
-
# Print first few rows of the processed DataFrame
|
84 |
-
print(df.head())
|
85 |
-
print(df.columns)
|
86 |
-
print(f"Features columns: {features.columns}")
|
87 |
-
print(f"Target unique values: {target.unique()}")
|
88 |
-
|
89 |
-
from sklearn.impute import SimpleImputer
|
90 |
-
from sklearn.preprocessing import StandardScaler
|
91 |
-
print(f"Missing values in features:\n{features.isnull().sum()}")
|
92 |
-
print(f"Missing values in target:\n{target.isnull().sum()}")
|
93 |
-
print(f"Infinites in features:\n{np.isinf(features).sum()}")
|
94 |
-
|
95 |
-
# Replace infinite values with NaN
|
96 |
-
features.replace([np.inf, -np.inf], np.nan, inplace=True)
|
97 |
-
|
98 |
-
# Handle missing values: Impute with the mean (for numerical features)
|
99 |
-
imputer = SimpleImputer(strategy='mean')
|
100 |
-
features_imputed = imputer.fit_transform(features)
|
101 |
-
|
102 |
-
# Normalize features to handle large values
|
103 |
-
scaler = StandardScaler()
|
104 |
-
features_scaled = scaler.fit_transform(features_imputed)
|
105 |
-
|
106 |
-
# Split data into training and testing sets
|
107 |
-
X_train, X_test, y_train, y_test = train_test_split(features_imputed, target, test_size=0.3, random_state=42, stratify=target)
|
108 |
-
|
109 |
-
# Initialize and train the Extra Trees model
|
110 |
-
model = ExtraTreesClassifier(n_estimators=100, random_state=42)
|
111 |
-
model.fit(X_train, y_train)
|
112 |
-
|
113 |
-
y_pred = model.predict(X_test)
|
114 |
-
print(classification_report(y_test, y_pred))
|
115 |
-
from sklearn.metrics import accuracy_score, classification_report
|
116 |
-
train_predictions = model.predict(X_train)
|
117 |
-
test_predictions = model.predict(X_test)
|
118 |
-
|
119 |
-
train_accuracy = accuracy_score(y_train, train_predictions)
|
120 |
-
test_accuracy = accuracy_score(y_test, test_predictions)
|
121 |
-
|
122 |
-
print(f"Training Accuracy: {train_accuracy:.4f}")
|
123 |
-
print(f"Testing Accuracy: {test_accuracy:.4f}")
|
124 |
-
|
125 |
-
print("Classification Report (Test Data):")
|
126 |
-
print(classification_report(y_test, test_predictions))
|
127 |
-
|
128 |
-
# Save the model and feature names
|
129 |
-
joblib.dump(model, 'extratrees.pkl')
|
130 |
-
joblib.dump(all_features.tolist(), 'featurenames.pkl')
|
131 |
-
import joblib
|
132 |
-
|
133 |
-
# Load the model and feature names
|
134 |
-
loaded_model = joblib.load('extratrees.pkl')
|
135 |
-
loaded_features = joblib.load('featurenames.pkl')
|
136 |
-
|
137 |
-
# Check if they are loaded successfully
|
138 |
-
print(f"Model Loaded: {loaded_model is not None}")
|
139 |
-
print(f"Features Loaded: {loaded_features is not None}")
|
140 |
-
|
141 |
-
# prompt: print different styles and new styles for the classification report\
|
142 |
-
import matplotlib.pyplot as plt
|
143 |
-
import seaborn as sns
|
144 |
-
from sklearn.metrics import classification_report
|
145 |
-
def plot_classification_report_styled(y_true, y_pred):
|
146 |
-
report = classification_report(y_true, y_pred, output_dict=True)
|
147 |
-
df_report = pd.DataFrame(report).transpose()
|
148 |
-
|
149 |
-
# Style the DataFrame with different colors and formatting
|
150 |
-
styled_report = df_report.style.background_gradient(cmap='viridis', axis=None) \
|
151 |
-
.highlight_max(color='lightgreen', axis=0) \
|
152 |
-
.highlight_min(color='lightcoral', axis=0) \
|
153 |
-
.format('{:.2f}')
|
154 |
-
|
155 |
-
# Display the styled report
|
156 |
-
display(styled_report)
|
157 |
-
|
158 |
-
|
159 |
-
# Use the new function to display a styled classification report
|
160 |
-
plot_classification_report_styled(y_test, y_pred)
|
161 |
-
|
162 |
-
|
163 |
-
# Alternative Styling using Seaborn and Matplotlib with customization
|
164 |
-
|
165 |
-
def plot_classification_report_seaborn_styled(y_true, y_pred):
|
166 |
-
report = classification_report(y_true, y_pred, output_dict=True)
|
167 |
-
df_report = pd.DataFrame(report).transpose()
|
168 |
-
plt.figure(figsize=(10, 6))
|
169 |
-
sns.heatmap(df_report[['precision', 'recall', 'f1-score']], annot=True, fmt=".2f", cmap="YlGnBu", linewidths=.5, annot_kws={"size": 12})
|
170 |
-
plt.title("Classification Report Heatmap", fontsize=16)
|
171 |
-
plt.xlabel("Metrics", fontsize=14)
|
172 |
-
plt.ylabel("Classes", fontsize=14)
|
173 |
-
plt.xticks(fontsize=12)
|
174 |
-
plt.yticks(fontsize=12)
|
175 |
-
plt.show()
|
176 |
-
|
177 |
-
plot_classification_report_seaborn_styled(y_test, y_pred)
|
178 |
-
|
179 |
import time
|
180 |
import subprocess
|
181 |
import pyshark
|
@@ -191,8 +13,8 @@ import requests
|
|
191 |
import gradio as gr
|
192 |
|
193 |
# Load the pre-trained model and feature names
|
194 |
-
model = joblib.load('
|
195 |
-
all_features = joblib.load('
|
196 |
|
197 |
# Modify the capture duration to a longer period
|
198 |
def capture_packets(url, capture_duration=30, capture_file="capture.pcap"):
|
@@ -352,36 +174,3 @@ iface = gr.Interface(
|
|
352 |
# Launch the interface
|
353 |
iface.launch(debug=True)
|
354 |
|
355 |
-
import matplotlib.pyplot as plt
|
356 |
-
import numpy as np
|
357 |
-
|
358 |
-
# Sample data extracted from captured packets
|
359 |
-
# These would come from the extracted packet features
|
360 |
-
tcp_counts = 20 # Number of TCP packets
|
361 |
-
udp_counts = 10 # Number of UDP packets
|
362 |
-
packet_sizes = [60, 150, 300, 450, 500, 700, 900, 1100, 1400, 1600] # Example packet sizes in bytes
|
363 |
-
timestamps = np.linspace(0, 30, len(packet_sizes)) # Sample timestamps over 30 seconds
|
364 |
-
|
365 |
-
# Create a new figure
|
366 |
-
plt.figure(figsize=(10, 6))
|
367 |
-
|
368 |
-
# Plot TCP and UDP packet counts in a bar chart
|
369 |
-
plt.subplot(2, 1, 1) # 2 rows, 1 column, first plot
|
370 |
-
plt.bar(['TCP', 'UDP'], [tcp_counts, udp_counts], color=['blue', 'orange'])
|
371 |
-
plt.title('TCP vs UDP Packet Counts')
|
372 |
-
plt.xlabel('Protocol')
|
373 |
-
plt.ylabel('Packet Count')
|
374 |
-
|
375 |
-
# Plot packet sizes over time
|
376 |
-
plt.subplot(2, 1, 2) # 2 rows, 1 column, second plot
|
377 |
-
plt.plot(timestamps, packet_sizes, marker='o', color='green')
|
378 |
-
plt.title('Packet Sizes over Time')
|
379 |
-
plt.xlabel('Time (s)')
|
380 |
-
plt.ylabel('Packet Size (bytes)')
|
381 |
-
|
382 |
-
# Adjust layout to prevent overlap
|
383 |
-
plt.tight_layout()
|
384 |
-
|
385 |
-
# Display the plots
|
386 |
-
plt.show()
|
387 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
import time
|
2 |
import subprocess
|
3 |
import pyshark
|
|
|
13 |
import gradio as gr
|
14 |
|
15 |
# Load the pre-trained model and feature names
|
16 |
+
model = joblib.load('extratrees.pkl')
|
17 |
+
all_features = joblib.load('featurenames.pkl')
|
18 |
|
19 |
# Modify the capture duration to a longer period
|
20 |
def capture_packets(url, capture_duration=30, capture_file="capture.pcap"):
|
|
|
174 |
# Launch the interface
|
175 |
iface.launch(debug=True)
|
176 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|